Measurement 46 (2013) 161170

Contents lists available at SciVerse ScienceDirect

Measurement
journal homepage: www.elsevier.com/locate/measurement

The measurement of thermal conductivity variation with temperature

for solid materials
S. Aksz a, E. ztrk b, N. Marasl c,
a
Nevsehir University, Faculty of Arts and Sciences, Department of Physics, 50300 Nevsehir, Turkey
b
Kocaeli University, Faculty of Arts and Sciences, Department of Physics, 41380 Kocaeli, Turkey
c
Erciyes University, Faculty of Sciences, Department of Physics, 38039 Kayseri, Turkey

a r t i c l e i n f o a b s t r a c t

Article history: An apparatus was designed to routinely measure the thermal conductivity variation with
Received 22 April 2012 temperature for solid materials. The apparatus was calibrated by measuring the thermal
Received in revised form 1 June 2012 conductivity variations with temperature for aluminum, zinc, tin and indium metals. The
Accepted 5 June 2012
variations of thermal conductivity with temperature for the Zn-[x] wt.% Sb alloys (x = 10,
Available online 26 June 2012
20, 30 and 40) were then measured by using the linear heat ow apparatus designed in
present work. From experimental results it can be concluded that the linear heat ow
Keywords:
apparatus can be used to measure thermal conductivity variation with temperature for
A. Metals
D. Thermal conductivity
multi component metallic alloys as well as pure metallic materials and for any kind of
D. Electrical conductivity alloys. Variations of electrical conductivity with temperature for the Zn-[x] wt.% Sb alloys
were determined from the WiedemannFranz (WF) equation by using the measured val-
ues of thermal conductivity. Dependencies of the thermal and electrical conductivities on
composition of Sb in the ZnSb alloys were also investigated. According to present exper-
imental results, the thermal conductivity and electrical conductivity for the Zn-[x] wt.% Sb
alloys decrease with increasing the temperature and the composition of Sb.
2012 Elsevier Ltd. All rights reserved.

1. Introduction
In the experimental determination of the thermal con-
ductivity of solids, a number of different methods of mea-
surement are required for different ranges of temperature
and for various classes of materials having different ranges
of thermal conductivity values. A particular method may
thus be preferable over-others for a given material and
temperature range.
The various methods for the measurement of thermal
conductivity fall into two categories: steady state and
non-steady state methods. In steady state methods of mea-
surement, the specimen is subjected to a temperature pro-
le that is time invariant; after equilibrium has been
reached; the thermal conductivity is determined directly
Corresponding author. Tel.: +90 352 437 49 01x33114; fax: +90 352
437 49 33.
E-mail address: marasli@erciyes.edu.tr (N. Marasl).
by measuring the rate of heat ow per unit area and the
temperature gradient. In non-steady state methods of
measurement, the temperature distribution in the speci-
men varies with time, and the measurement of the rate
of temperature change, which normally determines the
thermal diffusivity, replaces the measurement of the rate
of heat ow. The thermal conductivity is then calculated
from the thermal diffusivity with a further knowledge of
the density and specic heat of the materials [1].
Many attempts have been made to determine the ther-
mal conductivity values of solid and liquid phases in vari-
ous materials by using different methods. One of the
common techniques for measuring the thermal conductiv-
ity of solids is the longitudinal heat ow method. In the
longitudinal heat ow methods, the experimental arrange-
ment is so designed that the ow of heat is only in the axial
direction of a road specimen. Under steady-state condition
and assuming no radial heat loss or gain, thermal conduc-
tivity is determined by the following expression from

0263-2241/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.measurement.2012.06.003
162 S. Aksz et al. / Measurement 46 (2013) 161170
one-dimensional FourierBiot heat conduction equation
[1,2]:
Q DT
K 1
A DX
where K is the thermal conductivity of solid, Q is the rate of
heat ow or input power ows through to specimen, A is
the cross-sectional area of the specimen, DX = X2  X1 is
the distance between points of temperature measurements
for T1 and T2 and DT = T2  T1 is the temperature difference
between X2 and X1 points.
There are several different types of apparatus, all
employing the longitudinal heat ow methods. The longi-
tudinal heat ow methods are a steady state methods
and the classication is mainly based upon: absolute or
comparative methods. Rod method is one of the most com-
mon absolute methods and suitable for good conductors.
The specimen is in the form of a relatively rod so as to pro-
duce an appreciable temperature drop along the specimen
for precise measurement. In this method, a source of heat
at a constant temperature is supplied at the one end of
the rod and ows axially through the rod to other end,
where a heat sink at a lower constant temperature is lo-
cated. Radial heat loss or gain of rod should be negligible.
In order to determine the thermal conductivity from Eq.
(1), it is necessary to measure the rate of heat ow into
and/or out of the rod, cross-sectional area, the tempera-
tures of at least two points along the rod and the distance
between points of temperature measurements. For mea-
surements at lower (cryogenic) temperatures, radial heat
loss does not constitute a serious problem, and thermal
insulation and guard heaters are normally not necessary
[35]. For measurements at higher temperatures, radial
heat loss becomes serious problem because radiant heat
transfer increases rapidly with temperature. This method,
as used for measurements at higher temperatures has been
reviewed and discussed by Laubitz [6] and Flynn [7]. Mea-
surements of thermal conductivity at higher temperatures
have been made for different kind of materials [815].
Recently, we have embarked on research pertaining to
the measurement of thermal conductivity variation with
temperature of metals and multi-components of metallic
alloys. So attention is given to the thermal conductivity
and electrical conductivity at high temperature required
as input data in heat transfer and solidication simulators.
The aim of present work was to modify the rod method to
measure thermal conductivity variation with temperature.
For this purpose, a linear heat ow apparatus was designed
and calibrated by measuring the thermal conductivity var-
iation with temperature for the aluminum, zinc, tin and in-
dium metals. Then, the thermal conductivity variations
with temperature for the Zn-[x] wt.% Sb alloys (x = 10, 20,
30 and 40) were measured with the linear heat ow appa-
ratus designed in present work. The variations of electrical
conductivity with temperature for same alloys were also
determined from the WiedemannFranz (WF) equation
by using the measured values of thermal conductivity.
Dependencies of the thermal and electrical conductivities
on composition of Sb in the ZnSb alloys were also
investigated.
2. Experimental procedure
2.1. Experimental apparatus
As mentioned above, a linear heat ow apparatus was
designed to routinely measure the thermal conductivity
variation with temperature for solid materials. The linear
heat ow apparatus consists of hot stage, cold stage and
sample holder as shown in Fig. 1.
The hot-stage is comprised of two brass plates which are
resistively heated by NiCr wires, insulated in alumina tubes
and integrally threaded through the plates of the hot stage.
A total of 1000 mm of heater wire, 0.5 mm in diameter was
used in the hot-stage, providing a maximum power of
4500 W at 220 V AC. To maximize the thermal stability of
the hot-stage, a transformer was placed in the supply cir-
cuit, stepping the maximum current down to 4 A. A fully
proportional thermistor-based control system was imple-
mented, employing a control thermocouple within the
hot-stage. The temperature of the hot-stage was controlled
to an accuracy of 0.01 K with a Eurotherm 2604 type con-
troller. The hot stage can be operated up to 773 K degrees.
The cold-stage design is similar to that of the hot-stage.
However, cooling is achieved by using a Poly Science digi-
tal 9102 model heating/refrigerating circulating bath con-
taining an aqueous ethylene glycol solution. The
temperature of the circulating baths was kept constant at
278 K with an accuracy of 0.01 K.
To get linear temperature gradient into specimen, the
distance between the hot stage and cold stage was kept
typically 10 mm. Hot and cold stages were placed with a
space of 10 mm on an insulating plate and the insulating
plate was then vertically xed as the hot stage is at top
of the insulating plate and the cold stage is at bottom to
prevent convection effect on heat conduction into speci-
men and get a constant linear temperature gradient into
specimen as shown in Fig. 1.
Sample holder consists of two copper plates as shown in
Fig. 2. Two holes in 10 mm depth and 8 mm diameter were
drilled at cross sections of cold and hot copper plates to
place the specimen between the hot and cold stages and
get good heat conduction trough to specimen. At the same
time, a hole in 3 mm diameter through bottom cold copper
plate was also drilled to insert the measurements thermo-
couples into specimen. The ends of specimen were tightly
tted into holes at the cold and hot copper plates. Thermo-
couples were then placed into specimen by inserting ther-
mocouples through hole at the cold plate as shown in
Fig. 1. Then, top and bottom copper plates include the
specimen were placed together into hot and cold stages.
2.2. Specimen preparation and processing
A thin-walled graphite crucible, 12.0 mm OD  8.0 mm
ID  50 mm in length was made by drilling out a graphite
rod of 12 mm in diameter and 250 mm in length. A hole,
1.2 mm in diameter was drilled at the bottom of crucible
for thermocouples alumina tube as shown in Fig. 3.
ZnSb binary system shows an eutectic phase equilibria
in the range of compositions 057 wt.% Sb. The eutectic
 S. Aksz et al. / Measurement 46 (2013) 161170 163

Insulating plate Multimeter

-----AC V
Power input

Hot stage Controller thermocouple

Cold stage

-----AC I
Heating/ Multimeter
refrigerating
circulator

Fluid inlet
Fluid outlet

Measurement thermocouples

Data logger
Transformer
Temperature controller
Computer

Fig. 1. Block diagram of linear heat ow apparatus.

60 mm

Hot copper plate

12 mm

45 mm
8 mm 10 mm 8 mm
Cold copper plate
10 mm

3 mm
Measurement thermocouples 3 mm
4 mm

10 mm

Positions of measurement thermocouples into specimen

Fig. 2. Schematic illustration of copper plates and thermocouples positions into specimen.

melting temperature of ZnSb system is about 684 K. Be-
low the eutectic temperature, the ZnSb binary system
consists of solid Zn and solid e (Sb-43 wt.% Zn). Thus, com-
position of ZnSb alloys were chosen to be 10, 20, 30 and
40 wt.% Sb to see the affect of alloying Sb element on the
thermal conductivity of ZnSb alloy.
Tin, zinc, indium, aluminum and Zn-[x] wt.% Sb alloys
(x = 10, 20, 30 and 40) were separately melted in a vacuum
furnace by using metals of purity of 99.99%. After stirring,
the molten alloy was poured into a graphite crucible, held
in a specially constructed casting furnace at approximately
50 K above the melting point of the metals. The molten
metals were then directionally solidied from bottom to
top to ensure that the crucible was completely lled.
The solidied specimen was removed from the graphite
crucible and cut into lengths typically 30 mm. The 30 mm
164 S. Aksz et al. / Measurement 46 (2013) 161170
12 mm
8 mm
Graphite crucible
17 mm
Specimen
Alumina tube
23 mm
10 mm
Fig. 3. Schematic illustration of specimen.
length of specimen was placed into the linear heat ow
apparatus. Both ends of specimen (10 mm length) were in-
serted into holes at the hot and cold stages. Thus, a space of
10 mm between the hot stage and cold stage was adjusted
to get a linear temperature gradient into specimen and
prevent radial heat loss of rod. The temperatures at differ-
ent parts of the specimen were measured with mineral
insulated metal sheathed, 0.25 mm in diameter K type
(nickel chromiumnickel aluminum) three thermocouples,
xed within the specimen with a spacing of 23 mm as
shown in Fig. 2. Three thermocouples were placed with a
spacing of 23 mm and stuck together with silicone elasto-
mer glue. Before inserting the thermocouples into speci-
was measured with optical microscope to determine the
cross-sectional area (pr2). The temperatures of stationary
thermocouples were recorded with a Pico TC-08 data-log-
ger during the heat and the difference of temperature be-
tween two thermocouples, DT was read from the data
logger records. The distance between the two thermocou-
ples was measured from the photograph of the thermocou-
ples positions. In this experimental technique, the main
difculty was the measurement of heat ow rate into rod
specimen. To overcame this difculty, the input powers gi-
ven to experimental system (heater) for without specimen
(QWOS) and with specimen (QWS) were separately deter-
mined by measuring the voltage drops at the end of heater
and currents passing through the heater under the steady
state conditions. Heat ow rate into rod specimen (Q) for
each steady state condition was assumed to be the differ-
ence between the values of QWS and QWOS in present work.
After all desired temperature measurements had been
completed during the heat, the cooling procedure was
started in same steps down to room temperature.
The sample was then removed from the linear heat ow
apparatus and the longitudinal section of the specimen
was examined for porosity, cracks, and casting defects to
make sure that these would not introduce any errors to
the measurements.
Variations of thermal conductivity with temperature for
the Al, Zn, Sn, In and Zn-[x] wt.% Sb alloys (x = 10, 20, 30
and 40) were determined from Eq. (1) by using the mea-
sured values of A, Q = QWS  QWOS, DT and DX.
2.3. Experimental error analysis in the thermal conductivity
measurement
The estimated experimental error in the measurement
of K is the sum of the fractional uncertainty of the mea-
surements of heat ow rate, temperature difference, cross
sectional area of specimen and thermocouple positions
which can be expressed as
men, the photograph of thermocouples positions was          
taken to measure the distance between the thermocouples DK s  DQ  DT   DA DX 
       
 K   Q   DT   A   X  2
(DX) with optical microscope. All thermocouples were s
then inserted together into alumina tubes, 1.2 mm
OD  1.0 mm ID  20 mm in length and the ends of the
2.3.1. Fractional uncertainty in the heat ow rate
thermocouples were then connected to the measurement
measurement
unit which is consisted of data-logger and computer.
As mentioned above, the heat ow rate into specimen
The specimen was heated from one side by using a hot
was assumed to be the difference between the values of
stage in steps of 40 K up to 10 K below the melting temper-
QWS and QWOS for each steady state condition, i.e.
atures of the materials and the other side of specimen was
Q = QWS  QWOS. Typical data for measurements of heat
kept cool by using a cold stage to get linear temperature
ow rates are given in Table 1. Thus, the estimated exper-
gradient. As mentioned above the temperature of hot stage
imental error in the measurement of Q is the sum of the
was controlled to an accuracy of 0.01 K with a Eurotherm
fractional uncertainty in the measurements of QWS and
2604 type controller and the temperature of cold stage was
QWOS and can be expressed as
kept constant to an accuracy of 0.01 K with a Poly Science
     
digital 9102 model heating/refrigerating circulating bath. DQ  DQ WS  DQ WOS 
    3
The specimen was kept at steady state condition for at Q  Q  Q 
WS WOS
least 2 h for each temperature.
In order to determine the thermal conductivity from The input power is expressed as
Eq. (1), it is necessary to measure the rate of heat ow into
Q V I 4
the rod, cross-sectional area, the temperatures of at least
two points along the rod and the distance between points where V and I are the potential drop at the end of hot
of temperature measurements. The radius of rod specimen stage (heater) and the current passing through the heater,
 S. Aksz et al. / Measurement 46 (2013) 161170 165

Table 1
Typical experimental data to determine heat ow rate into aluminum rod specimen at different temperatures.

Temperature (K) IWOS (A) VWOS (V) QWOS (W) IWS (A) VWS (V) QWS (W) Q = QWS  QWOS (W)
313 0.77 7.59 5.87 0.9 9.1 8.3 2.4
353 1.39 13.04 18.12 1.8 17.6 31.2 13.1
393 1.95 19.34 37.71 2.3 23.6 54.9 17.2
433 2.35 23.50 55.23 2.9 28.5 82.1 26.9
473 2.88 29.62 85.32 3.2 35.0 112.6 27.3
513 3.28 33.35 109.46 3.7 37.9 140.2 30.8
553 3.56 36.79 130.91 4.1 41.8 169.7 38.8
593 3.87 39.75 153.66 4.4 45.8 202.7 49.1
633 4.28 44.05 188.46 4.8 49.6 237.1 48.7
673 4.66 47.86 222.77 5.2 53.9 277.7 54.9

IWOS: Current passing through to heater without specimen; IWS: Current passing through to heater with specimen; VWOS: Potential drop at the ends of heater
without specimen; VWS: Potential drop at the ends of heater with specimen; QWOS: Input power given to heater without specimen; QWS: Input power given
to heater with specimen; Q: Heat ow rate through to specimen.

respectively under the steady state condition. The fractional ple readings (DT) at the same points of the specimen must
uncertainty in the power measurement can be expressed as be known or measured to determine the uncertainty of
     
DQ  DV  DI temperature measurement. To determine the difference
    5
Q  V  I of the thermocouple readings, the thermocouples were cal-
Potential drop and the current were measured with a
HewlettPackard 34401-A multimeter. The variations in
the current reading (DI) were 0.02 A at different temper-
atures. Typical data to determine heat ow rates through
to aluminum rod at different temperatures are given in
Table 1. As can be seen from Table 1, the currents passing
through to specimen were in the range of 0.95.2 A at dif-
ferent temperatures. The uncertainties in the current mea-
surements are in the range of 0.32.2%. The estimated error
in the current measurement is about 4.4%.
The variations in the potential drop readings (DV) were
0.03 V at different temperatures. As can be seen from Ta-
ble 1, the potential drops at the end of hot stage were in
the range of 9.153.9 V. The uncertainties in the potential
drop measurements are in the range of 0.050.3% at differ-
ibrated by detecting the melting point of a metallic mate-
rial. The difference between the two thermocouple
readings (DT) were in the range of 0.030.2 K at different
temperatures [16]. As can be seen from Table 2, the tem-
perature differences into the specimens (DT = T1  T2)
were in the range of 1.230.6 K at different temperatures.
The uncertainties in the temperature measurement into
the specimen are in the range of 0.62.5% at different tem-
peratures. The estimated error in the temperature mea-
surement into the specimen is about 2.5%.
2.3.3. The fractional uncertainty in the measurement of cross-
sectional area (A) and the xed distances (X1, X2)
The cross sectional area of specimen (A) is pr2 and the
fractional uncertainty in the cross sectional surface mea-
surement is expressed as
ent temperatures. The estimated error in the potential drop    
DA  
measurement is about 0.6%.   2Dr 6
Thus, the total fractional uncertainty in heat ow rate
measurement is about 5%.
2.3.2. Fractional uncertainty in the measurement of
temperature difference between two thermocouples, DT=T1
 T2 at the setting temperature
The temperature of the specimen was measured with K
type thermocouples. The difference of the two thermocou-
A r
where r is the radius of specimen. The radius of specimen is
measured to be 4000 lm by using an optical microscope
with an accuracy of 10 lm. Thus, the uncertainty in the
measurement of radius or cross sectional area of specimen
is less than 0.5%.
The distance between two thermocouple positions
(DX = X2  X1) was also measured by using an optical

Table 2
Experimental data in the measurements of thermal conductivity variations with temperature for the aluminum, zinc, indium and tin.

Temperature (K) Aluminum (Al) Zinc (Zn) Indium (In) Tin (Sn)
K (W/Km) DT (K) Q (W) K (W/Km) DT (K) Q (W) K (W/Km) DT (K) Q (W) K (W/Km) DT (K) Q (W)
313 238.0 1.2 2.4 117.6 1.7 1.2 82.7 2.4 3.3 66.9 4.7 5.3
353 234.3 6.7 13.1 113.3 2.8 2.1 82.0 4.8 6.6 65.0 8.9 9.7
393 232.0 8.9 17.2 111.7 5.1 5.6 80.8 6.1 8.3 63.0 13.6 14.4
433 230.5 13.9 26.9 110.3 8.5 8.3 80.0 8.4 11.2 61.2 19.4 19.8
473 228.0 14.3 27.3 105.6 9.9 10.1 80.0 10.0 13.4 59.4 24.9 24.8
513 224.1 16.4 30.8 104.0 24.6 32.5 78.9 12.6 16.6
553 221.2 20.9 38.80 100.8 33.5 47.1
593 218.6 26.8 49.06 99.5 40.4 56.1
633 216.7 26.8 48.67 97.6 45.6 60.2
673 214.3 30.6 54.92

K: Thermal conductivity of specimen (W/Km); DT: Temperature difference into specimen (K); Q: Heat ow rate into specimen (W).
166 S. Aksz et al. / Measurement 46 (2013) 161170

microscope with an accuracy of 10 lm. The measured va-
lue of DX = X2  X1 is about 3000 lm. The fractional uncer-
tainty in the measurement of the xed distance is about
0.3%.
Therefore the total fractional uncertainty in the mea-
surements of thermal conductivity in present work is
about 9%.
2.4. Determination of the thermal temperature coefcient
For a given composition, the dependence of the thermal
conductivity of the solid phase on temperature is linear
and the thermal temperature coefcient, aTTC is expressed
as [17,18]
K  K0 1 DK
aTTC
K 0 T  T 0 K 0 DT
where K is the thermal conductivity at the temperature of
T, K0 is the thermal conductivity at the room temperature,
T0 = 300 K and aTTC is the thermal temperature coefcient.
This means that the thermal temperature coefcient, aTTC
can be obtained from the graph of thermal conductivity
variations with temperature and also given in Table 3.
2.5. Determination of electrical conductivity
Heat in solid is conducted by various carriers: electrons,
lattice waves or phonons, magnetic excitations, and, in
some cases, electromagnetic radiation. The total thermal
conductivity is additively composed of contributions from
each type of carrier. The principal carriers of heat in metals
are electron and lattice waves, leading to an overall ther-
mal conductivity
K Ke Kg
where Ke is the electronic component and Kg is the lattice
component.
Generally Kg of metals, alloys and semimetals is of mag-
nitude comparable to the lattice thermal conductivity of
insulators of corresponding elastic properties, except at
low temperature where phononelectron interaction re-
duces Kg in metals. The relative importance of Ke and Kg
thus depends on the magnitude of Kg. The electronic com-
ponent often parallels the electrical conductivity and the
electrical conductivity is highest in pure metals, reduced
in the case of alloys and, even lower in semimetals and
semiconductors. Electrical conductivity is a measure of a
materials ability to conduct an electric current and is
one of the primary physical properties of materials such
as thermal conductivity, specic heat and thermal expan-
sion. The relationship between the thermal conductivity
and the electrical conductivity of metals is established by
the WiedemannFranz equation [19], which is based on
the fact that heat and electrical transport both involve
the free electrons in the metal as:
Ke
LT 9
r
where L is the constant of proportionality, which is called
the Lorenz number. The value of L is 2.45  108 WX/K2.
The variation of electrical conductivity (r) with tempera-
ture can be determined from the WiedemannFranz equa-
7
tion by using the measured values of K.
2.6. Determination of electrical temperature coefcient
The electrical conductivity is strongly dependent on
temperature. In metals, electrical conductivity decreases
with increasing temperature, whereas in semiconductors,
electrical conductivity increases with increasing tempera-
ture. The dependence of electrical conductivity on temper-
ature is often expressed as a slope in the electrical
conductivity versus temperature graph and can be given
as:
r  r0 1 Dr
aETC 10
r0 T  T 0 r0 DT
where r is the electrical conductivity at the temperature
(T), r0 is the electrical conductivity at the room tempera-
ture, T0 = 300 K and aETC is the electrical temperature
8 coefcient.
3. Results and discussions
3.1. Calibration of linear heat ow apparatus
As mentioned above, the rod method is one of the most
common absolute method and suitable for good conduc-
tors. In this method, a source of heat at a constant temper-
ature is supplied at the one end of the rod and ows axially
through the rod to other end, where a heat sink at a lower
constant temperature is located. Radial heat loss or gain of

Table 3
Some thermal and electrical properties of tin, zinc, indium, aluminum, and Zn-[x] wt.% Sb alloys (x = 10, 20, 30 and 40) at their melting temperature.

Metals Melting Thermal conductivity Thermal temperature Electrical Electrical

temperature (K) K (W/Km) coefcient conductivity temperature
aTTC (K1) r  108 (1/Xm) coefcient aETC (K1)
Tin (Sn) 505 57.80 0.000692 0.0413 0.002554
Zinc (Zn) 692 93.07 0.000524 0.0406 0.001874
Indium (In) 429 78.27 0.000444 0.0718 0.002710
Aluminum (Al) 933 237.35 0.000273 0.0115 0.001518
Zn-10 wt.% Sb 682 55.68 0.000477 0.0181 0.002073
Zn-20 wt.% Sb 682 42.93 0.000433 0.0139 0.002068
Zn-30 wt.% Sb 682 30.86 0.000767 0.0080 0.002238
Zn-40 wt.% Sb 682 21.68 0.000881 0.0049 0.002308
 S. Aksz et al. / Measurement 46 (2013) 161170 167

rod should be negligible. In order to determine the thermal In addition, the thermal temperature coefcient of the Zn,
conductivity with rod method, it is necessary to measure Sn, In and Al were also found to be 0.000524, 0.000692,
the rate of heat ow into and/or out of the rod, cross sec- 0.0004444 and 0.000273 K1, respectively from Fig. 4.
tional area, the temperatures of at least two points along A comparison of the variation of thermal conductivity, K
the rod and the distance between points of temperature with temperature for the Zn, Sn, In and Al measured in
measurements. Main difculty in the rod method was the present work with the variation of thermal conductivity,
measurement of heat ow rate into and/or out of the rod. K with temperature for same metals measured in previous
In present work, this problem was overcome by measuring work [17] is given in Fig. 4. As can be seen from Fig. 4, the
the input powers given to experimental system (heater) for lines of thermal conductivity variation with temperature
with specimen (QWS) and without specimen (QWOS) at each measured in present work are slightly above than the lines
steady state conditions. of thermal conductivity variations with temperature for
For condence of experimental technique, the calibra- the same metals measured in previous works [17] and re-
tion of experimental apparatus must be made. For this pur- sults are in a good agreement with the previous works in
pose, Al, Zn, Sn, and In metals were chosen as test materials the range of experimental error. This means that the pres-
because of their thermal conductivity variations with tem- ent experimental technique is capable to easily and sensi-
perature are well known [17]. Same sized specimens for tively determine the variation of thermal conductivity with
Zn, Sn, In and Al were heated from one side by using a temperature for solid materials.
hot stage in steps of 40 K up to 10 K below the melting
temperature of the metals. The other side of specimens 3.2. Dependency of the thermal conductivity on the
was kept cool by using a cold stage to get linear tempera- temperature and composition of Sb in the Zn-[x] wt.% Sb alloys
ture gradient during the annealing period. The specimen
was kept at steady state condition for at least 2 h for each The eutectic melting temperature of Zn-[x] wt.% Sb bin-
temperature. The temperature difference between two ary system is about 684 K [20]. Zn-[x] wt.% Sb eutectic sys-
thermocouples (DT) for each temperature (steady state tem consists of solid e (Sb3Zn4) and solid Zn according to
condition) was read from the data logger record. The ther- the composition of Sb up to 57 wt.% Sb below the eutectic
mal conductivity of specimens for each steady state condi- melting temperature [20]. In present work, the variations
tion were obtained from Eq. (1) by using the measured of thermal conductivity with temperature for the Zn-
values of A, Q, DT and DX. [x] wt.% Sb alloys (x = 10, 20, 30 and 40) were also mea-
The variations of thermal conductivity with tempera- sured with the linear heat ow apparatus designed in pres-
ture for the Zn, Sn, In and Al were given in Table 2 and plot- ent work. Experimental results are given in Table 4 and
ted in Fig. 4. The thermal conductivity of Zn, Sn, In and Al at plotted in Fig. 5.
their melting temperature were found to be 93.07, 57.80, A comparison of the values of K for the Zn-[x] wt.% Sb al-
78.27 and 237.35 W/km, respectively by extrapolation the loys measured in present work with the values of K for Sb
thermal conductivity curves to the melting temperatures. [17] and Sn [17] are also given in Fig. 5. As can be seen from
Thermal Conductivity (W/Km)
Thermal Conductivity (W/Km)

250 70
Al [17] 68 Sn [17]
Al [PW] Sn [PW]
240 66
64
230 62
60
220 58
56
210 54
200 300 400 500 600 700 280 320 360 400 440 480 520
Temperature (K) Temperature (K)
Thermal Conductivity (W/Km)

Thermal Conductivity (W/Km)

84
115
In [17] Zn [17]
82 In [PW]
110 Zn [PW]

80
105
78
100
76
95
74
300 320 340 360 380 400 420 300 350 400 450 500 550 600 650
Temperature (K) Temperature (K)

Fig. 4. Variations of thermal conductivity with temperature for the Al, Sn, In and Zn metals.
168 S. Aksz et al. / Measurement 46 (2013) 161170

Table 4
Thermal conductivity variations with temperature for Zn-[x] wt.% Sb alloys (x = 10, 20, 30 and 40).

Temperature (K) Zn-10 wt.% Sb Zn-20 wt.% Sb Zn-30 wt.% Sb Zn-40 wt.% Sb
K (W/Km) DT (K) Q (W) K (W/Km) DT (K) Q (W) K (W/Km) DT (K) Q (W) K (W/Km) DT (K) Q (W)
313 67.19 3.04 3.43 50.93 4.38 3.74 43.07 3.27 2.36 31.92 5.22 2.79
353 66.15 7.27 8.06 50.69 8.94 7.59 41.77 7.10 4.96 30.82 12.90 6.66
393 65.61 11.11 12.21 49.54 13.41 11.12 40.34 11.39 7.69 29.94 16.86 8.46
433 64.43 14.70 15.86 48.78 16.43 13.42 39.78 14.50 9.66 28.97 23.18 11.25
473 63.17 23.13 24.46 48.05 19.17 15.42 38.20 18.08 11.56 27.13 29.98 13.62
513 60.47 23.86 24.16 47.79 22.87 18.31 37.91 21.02 13.34 26.52 36.93 16.40
553 59.57 28.08 28.01 45.11 26.88 20.31 33.81 27.30 15.46 25.56 44.10 18.88

K: Thermal conductivity of specimen (W/Km); DT: Temperature difference into specimen (K); Q: Heat ow rate into specimen (W).

Fig. 5, the value of K decreases with increasing the temper-

70
ature. The measured lines of thermal conductivity varia-
313
tions with temperature for the Zn-[x] wt.% Sb alloys lie

Thermal Conductivity (W/Km)

353
between the lines of thermal conductivity variation with 393
60 433
temperature for Sb [17] and Zn [17]. As mentioned above, 473
513
the Zn-[x] wt.% Sb eutectic system consists co-exit two dif- 553
50
ferent solid phases as solid e (Sb3Zn4) and solid Zn below
the eutectic melting temperature. If the composition of Sb
in the ZnSb increases the mole fraction of intermetallic 40
phase (Sb3Zn4) increases. Thus, the lines of thermal conduc-
tivity variation with temperature for that alloys should be
below the line of thermal conductivity variation with tem- 30
perature for zinc. The values of K for the Zn-[x] wt.% Sb al-
loys (x = 10, 20, 30 and 40) at their melting temperature 20
were obtained to be 55.68, 42.93, 30.86, and 21.68 W/km, 5 10 15 20 25 30 35 40 45
respectively by extrapolating the thermal conductivity Composition (wt.%Sb)
curves to their melting temperature as shown in Fig. 5. In
Fig. 6. The variations of thermal conductivity with composition of Sb in
addition, the thermal temperature coefcient of the Zn-
the Zn-[x] wt.% Sb alloys at different temperatures.
[x] wt.% Sb alloys (x = 10, 20, 30 and 40) were also found
to be, 0.000477, 0.000433, 0.000767 and 0.000881 K1,
respectively from Fig. 5. 3.3. Dependency of the electrical conductivity on the
The thermal conductivity variations with composition of temperature and composition of Sb
Sb in the Zn-[x] wt.% Sb alloys (x = 10, 20, 30 and 40) were
plotted in Fig. 6 at different temperatures (313, 353, 393, The experimental error in the measurements of thermal
433, 473, 513 and 553 K). As can be seen from Fig. 6, the conductivity variation with temperature is about 9% by
thermal conductivity of Zn-[x] wt.% Sb alloys exponentially using the radial heat ow apparatus. Experimental error
decrease with increasing the composition of Sb. in the measurement of electrical conductivity is generally
higher than 9%. Thus, the variations of electrical conductiv-
ity with temperature determined from the Wiedemann
Franz (WF) equation for the Zn-[x] wt.% Sb alloys are plot-
ted in Fig. 7. There is no information about on the electrical
120
Thermal Conductivity (W/Km)

conductivity variations with temperature for the Zn-

Sb
100 Zn [x] wt.% Sb alloys. Thus, a comparison of the variations of
Zn-10 wt.%Sb
Zn-20 wt.%Sb
electrical conductivity with temperature determined from
80 Zn-30 wt.%Sb the WF equation for the Zn-[x] wt.% Sb alloys with the
Zn-40 wt.%Sb variations of electrical conductivity with temperature for
60 Sb and Zn measured in previous work [21] is also given
in Fig. 7. As can be seen from Fig. 7, the value of r for
40 the Zn-[x] wt.% Sb alloys decreases with increasing the
temperature and the calculated lines of r from the WF
20 equation for the Zn-[x] wt.% Sb alloys lie between the lines
of electrical conductivity variations with temperature for
0 Sb [21] and Zn [21].
200 300 400 500 600 700
The values of r for the Zn-[x] wt.% Sb alloys (x = 10, 20, 30
Temperature (K)
and 40) at their melting temperatures were obtained to be
Fig. 5. Variations of thermal conductivity with temperature for the Zn 18.1  103, 13.9  103, 8  103, and 4.9  103 (Xm)1,
[17], Sb [17] and Zn-[x] wt.% Sb alloys. respectively by extrapolating the electrical conductivity
 S. Aksz et al. / Measurement 46 (2013) 161170
Fig. 7. Variations of electrical conductivity with temperature for the Zn
[21], Sb [21] and Zn-[x] wt.% Sb alloys.
Fig. 8. The variations of electrical conductivity with composition of Sb in
the Zn-[x] wt.% Sb alloys at different temperatures.
curves to their melting temperature. The electrical tempera-
ture coefcients for the Zn-[x] wt.% Sb alloys (x = 10, 20, 30
and 40) were found to be 0.002073, 0.002068, 0.002238
and 0.002308 K1, respectively from Fig. 7 and also given
in Table 4.
The variations of electrical conductivity with composi-
tions of Sb in the Zn-[x] wt.% Sb alloys determined from
the WF equation were plotted in Fig. 8 at different tem-
peratures (313, 353, 393, 433, 473, 513 and 553 K). As
can be seen from Fig. 8, the electrical conductivity of Zn-
[x] wt.% Sb alloys linearly decreases with increasing the
composition of Sb.
4. Conclusions
The thermal conductivity (K) and electrical conductivity
(r) are the main fundamental properties of materials such
as density, melting point, entropy, enthalpy, resistance,
and crystal structure parameters. Although the values of
K and r for pure metallic materials were obtained theoret-
ically and experimentally, there are not enough informa-
tion and data available about the thermal and electrical
conductivity of metallic alloys. The values of K and r for
169
alloys change, as in pure materials, not only with temper-
ature but also it changes by compositions of alloys.
The measurement of thermal conductivity variation
with temperature for solid is difcult and very expensive.
To overcome these problems a linear heat ow apparatus
was designed to routinely measure the thermal conductiv-
ity variation with temperature for solids. To get a linear
temperature gradient into specimen, one side of specimens
was heated by using a hot stage and the other side of spec-
imens was kept cool by using a cold stage. The main dif-
culty in the experimental technique was the measurement
of heat ow rate into the rod specimen. This problem was
overcome by measuring the input powers given to experi-
mental system (heater) for with specimen (QWS) and with-
out specimen (QWOS) at each steady state conditions.
The variations of thermal conductivity with tempera-
ture were measured for the Zn, Sn, In and Al by using the
linear heat ow apparatus and the results were compared
with the known values of thermal conductivity variation
with temperature for the same metals measured in previ-
ous work [17]. From the comparison of present experimen-
tal results with previous results, it can be concluded that
the results are in a good agreement with the previous
works in the range of experimental error. The present
experimental technique has the advantages of simpler
apparatus, easier specimen fabrication and easier opera-
tion and can be used to determine the thermal conductiv-
ity variation with temperature for metals as well as
metallic and organic alloys. Present experimental tech-
nique might also be used to determine the thermal con-
ductivity variation with temperature for liquid materials.
Dependency of thermal conductivity of the Zn-[x] wt.%
Sb alloys on the temperature and composition of Sb were
investigated by using a linear heat ow apparatus designed
in present work. The thermal conductivities of the Zn-
[x] wt.% Sb alloys decrease with increasing the tempera-
ture and composition of Sb.
The electrical conductivity variations with temperature
for the Zn-[x] wt.% Sb alloys were determined from the
WiedemannFranz (WF) equation by using the measured
values of thermal conductivity for same alloys. The electri-
cal conductivities of Zn-[x] wt.% Sb alloys decrease with
increasing the temperature and composition of Sb.
Acknowledgements
This project was supported by the Erciyes University
Scientic Research Project Unit under Contract No.: FBA-
10-3297 and the Nevsehir University Scientic Research
Project Unit under Contract No.: 2010/15. Authors would
like to thank to the Erciyes University Scientic Research
Project Unit and Nevsehir University Scientic Research
Project Unit for their nancial support.
References
[1] J.B. Biot, Traite de Physique, vol. 669, Paris, 1816.
[2] J.B.J. Fourier, The Analytical Theory of Heat, Dover Publication, New
York, 1955.
[3] G.K. White, Experimental Techniques in Low-Temperature Physics,
Oxford University Press, Oxford, 1959.
170 S. Aksz et al. / Measurement 46 (2013) 161170
[4] G.K. White, Measurement of Solid Conductors at Low Temperatures
in Thermal Conductivity, vol. 1, Academic Press, London, 1969
(Chapter 2).
[5] C.H. Lees, The effects of temperature and pressure on the thermal
conductivities of solids part II. The effects of low temperatures on the
thermal and electrical conductivities of certain approximately pure
metals and alloys, Phil. Trans. Roy. Soc. London A208 (1908) 381
443.
[6] M.J. Laubitz, Measurement of the Thermal Conductivity of Solids at
High Temperatures by Using Steady-State Linear and Quasi-Linear
Heat Flow in Thermal conductivity (R.P. Type ed.), vol. 1, Academic
Press, London, 1969 (Chapter 3).
[7] D.R. Flynn, Thermal Conductivity of Ceramics in Mechanical and
Thermal Properties of Ceramics (J.B. Wachtman Jr., ed.), vol. 303, NBS
Spec. Publ., 1969, pp. 63123.
[8] A. Berget, Measurement of the thermal conductivity of mercury of its
absolute value, Compt. Rend. 105 (1887) 224227.
[9] A. Berget, Measurement of the thermal conductivity of mercury and
certain metals, J. Phys. (Paris) 2 (7) (1888) 503518.
[10] The Physical Societys Exhibition No. III, Engineer, 1935, pp. 15968.
[11] L.D. Armstrong, T.M. Dauphinee, Thermal conductivity of metals at
high temperatures. I. Description of the apparatus and measurement
on iron, Can. J. Res. A 25 (1947) 357374.
[12] D.A. Ditmars, D.C. Ginnings, Thermal conductivity of beryllium oxide
from 40 to 750 C, Res. Nat. Bur. Stand. 59 (2) (1957) 9399.
[13] K. Honda, T. Simidu, On the thermal and electrical conductivities of
carbon steels at high temperatures, Sci. Rep. Tohoku Univ. 1 (6)
(1917) 219233.
[14] F.H. Schoeld, The thermal and electrical conductivities of some
pure metals, Proc. Roy. Soc. (London) A107 (1925) 206227.
[15] R.W. Powell, The thermal and electrical conductivities of some pure
metals, Proc. Roy. Soc. (London) A107 (1925) 206227.
[16] S. Akbulut, Y. Ocak, K. Keslioglu, N. Marasl, Thermal conductivities
of solid and liquid phases for Neopentylglycol,
Aminomethylpropanediol and their binary alloy, J. Phys. Chem.
Solids 70 (2009) 72.
[17] Y.S. Touloukian, R.W. Powell, C.Y. Ho, P.G. Klemens, Thermal
Conductivity Metallic Elements and Alloys, IFI/Plenum, New York,
1970, pp. 17a. 49, 149, 185, 408, 498.
[18] H.L. Callendar, J.T. Nicolson, Experiments on the condensation of
steam. Part I. A new apparatus for studying the rate of condensation
of steam on a metal surface at different temperatures and pressures,
Brt. Assoc. Adv. Sci., Rept. Ann. Meeting 22 (1897) 418.
[19] C. Kittel, Introduction to Solid State Physics, 6th ed., John Wiley and
Sons, New York, 1965. p. 150.
[20] T.B. Massalski, J.L. Murray, L.H. Benett, H. Baker, Binary Alloy Phase
Diagrams, vol. 1, 2nd ed., ASM International, Materials Park, OH,
1990.
[21] E.A. Brandes, G.B. Brook, Smithells Metals Reference Book, vol. 14,
Elsevier, 1998.